Indoor 3D REM Design using Ordinary Kriging Interpolation

This study examines three-dimensional (3D) and two-dimensional (2D) interpolation approaches for indoor 3D radio environment map (REM), on received signal strength (RSS) data of ultrahigh frequency (UHF) active television (TV) channel, to explore indoor TV white space (TVWS) volumes. It could serve as a valuable resource for cognitive radio operations during unlicensed secondary access indoors. A significant portion of data traffic originates indoors. Altitude considerations or realistic view of indoor radio environment map could improve the detection of TVWS volumes, thereby boosting channel data rates and throughput, which are crucial in secondary access. The measured 3D RSS dataset assisted spatial interpolation, exponential and spherical semivariogram (SV) models based 3D and 2D ordinary kriging (OK), results evaluation has been done on two metrics - computation time (CT) and root mean square error (RMSE). Performance comparison of the spherical and exponential SV model for 3D and 2D OK, reveals that the spherical model outperforms the exponential model to a moderate extent for 3D-OK, while the exponential model marginally outdoes the spherical model in 2D-OK. Study finds that 3D-OK yields lower RMSE values compared to 2D-OK for spherical and exponential SV. The analysis indicates that 3D-OK achieves 17.41% and 13.98% higher accuracy using the spherical and exponential SV than 2D-OK. The CT of 3D-OK is 1.09 times faster than the 2D-OK. The best-tracked TVWS volume is 18200m³, 7.624% of the real building volume. The findings suggest considerable likelihood of accessing TVWS based on the RSS status of 3D indoor active TV channels.